I've also noticed that sometimes looking at in trade pairs and waiting for a first minor reversal back towards the mean is a good time to jump. These sometimes revert to mean within one to three days. Right now I'm also seeing a lot of ratio charts trending and you are playing against the trend. My biggest problem is trying to pick one chart to act on out of many that look the same. So far I've closed two for two for profit and one in trade under water. I'll keep plugging.
This would be a very good approximation. Helps a lot for quick mental calculation, but will not give the exact result in a spreadsheet. If the ratio is the Long stock divided by the Short stock, I think that for the accurate % profit one should multiply this by the ratio of the Short stock closing price to its entry price. In formulae: Long stock entry price L1 , exit price L2 Short stock entry price S1 , exit price S2 Entry ratio R1 = L1 / S1 Exit ratio R2 = L2 / S2 % gain is 100*[L2/L1 - S2/S1] / 2 = 100* [ (L2/S2) / (L1/S1) - 1 ] *(S2/S1) /2 = 100*[ (R2/R1) - 1 ] * (S2 / S1) Here 100*[ (R2/R1) - 1 ] is the percent change in the ratios, (S2/S1) is [1+100*% change of short stock price] .
Yes some pair traders do this, wait for reversion to start before entering, Ive always found that entering at the extreme's to offer the best risk vs reward profile, yes waiting for a reversal may have a higher win rate but at the expense of a lower win loss rate, your payoff is smaller. I wouldn't worry too much about picking the right chart, just develop a simple method and stick to it, staying disciplined is the biggest challenge in successful trading, many traders equate more work, harder work, longer work with becoming successful, the opposite is true, try and see how little you can spend trading per day, I aim for 30 mins or less, but for those 30 mins im at 110% concentration, ive had my coffee and red bull, chewing gum whilst looking for and executing trades, then i log out for the day, staying glued to the trading screen all day isn't good for your health, day trading symptoms aren't pretty, don't constantly look at your open PnL otherwise your emotions will start mimicking the charts we trade, less really is more. this is why you often hear successful traders say trading is boring, the novelty and romance wears off pretty quickly, don't get me wrong trading is a great job to those that can do it, but its not about executing a trade and pretending there might be a chance of loss before closing it for a profit every time, you gotta have the guts to stick it out and hold through the draw downs and soft patches. focus on keeping your emotions in check, and the charts will take care of themselves.
closed 1 trade; Sold SII @ 24.19 Covered DRQ @ 24.12 2 new trades; Long HP @ 22.31 Short UNT @ 26.45 Long WWW @ 17.18 Short NKE @ 47.20
HI All, pair trading newbie here, first post (so be gentle ) I have recently purchased PairTrade Finder and I'm in the process of loading up various pairs in industry groups. Now being based in Australia, I've come to the conclusion our local ASX is too small a market for pair trading, so it looks like NYSE/NASDAQ and/or LSE If I'm doing things correctly in PairTrader, I won't have a shortage of pairs to trade, however, the big problem I have is brokerage. If I use a local provider to trade CFD's on the NYSE, the brokerage is $10 USD for up to $10K trade, then increases over this lot size. I know you guys don't pay much at all in comparison. So for any given round trip trade per pair I'm up for at least $40 USD, which is about $60 AUD. I guess I could go and open an account with say IB, but it's not so straight forward for reason's I won't get into right now. So, if I'm not going to get eaten alive by brokerage, I can only think of 2 things. 1/ Increase my leverage (which isn't something I like doing until I have more thorough knowledge of the system characteristics ) or 2/ increase the average profit per trade Now for 2/ I was thinking of changing the default "stretch" from 2.0 to say 3.0 I figure this will give me less trades per pair, but the trades I get should be bigger as a consequence of moving the std dev out to 3. Reducing the number of trades I don't think will hurt too much as it seems there are numerous (>150) pairs for NYSE and I haven't even started on the LSE yet. Just like to know what people's thoughts are about this situation and my possible solution. It's a great thread and I've enjoyed reading most of it TIA Ivan
I'm new to this too, so take this with a grain of salt. In theory a larger Std. Dev. should produce better results, but you have to look harder at why the spread has gotten so large. Make sure it isn't news or earnings related and possibly check the volume during this time to see if maybe there is heavy buying or selling going on. Either of those situations I think would make the signal unreliable. I'm sure there are other things to think of too.
Jonny, Are you aware that SPG will be going ex-dividend very soon? I think it's 90 cents payable 90% in stock and 10% cash. Because you are short you will probably have to pay this dividend. This is the first time they are paying this dividend mostly in stock. In the past it was strictly cash. I wonder if this change has an significant impact on the correlation data. Don
Looking at the ASX I think you will be able to find plenty of good pairs to trade. When backtesting, filter your pairs for above $300 average profit per trade, therefore your brokerage is well covered, but yes having low commissions is vital to any trading success, focus on good execution aswell, try and trade inside the spread, buy on the bid and sell on the ask, this small edge over time really adds up. You can change the universal stretch to 3.00 and backtest that setting, so you can see what the average profit per trade is and if its better. To reduce your commissions relative to profits try trading more volatile pairs with less size.